For Christmas I decided to make a set of matching wooden planes for myself and my dad. Check out some snapshots of the process and the final product.
Moving on from the mechanical design, I could finally jump into developing the electronics to go inside the Tron Identity Disc. If you haven’t already, go check out the mechanical design here.
With the mechanical design complete I needed to design the electronics to make the Tron Identity Disc actually cool and do something other than just sit there. I would need to design a few different circuit boards to hold different sections and perform different tasks. Each of these was already determined from the mechanical design. I had already identified and chosen some parts during the CAD process but the schematic can be completed. Following this the circuit board layout. As well as the not planned stuff. Things such as continuous back and froth between CAD packages and models, and datasheets and schematics.
Picking the parts was a mixed difficulty task, some really easy and knew straight away others took a bit of research, digging and asking around. Allocating footprints was easy as I chose parts I could solder by hand making the smallest package 0603 and a few other may just take a bit more patience to get into place.
- SK6812 Addressable RGB LED, in both 5050 package (Light Arc)and side mounted (Main PCB)
- STM32, more specifically aiming for the
- Batteries, 6x
- 9DoF IMU, LSM9DS1
- Magnetic field Sensors,
- Power Regulators, booster and switches to control power around the boards
- Passive components like resistors and capacitors in 0805 or 0603 packages, as that’s the smallest I can easily solder by hand
- and everything else required to work
The schematic was broken down into key sections to make it easier to manage. This was applied to both circuit boards being developed. Each major section of the design was put into it’s own designated schematic to make it easier t o find sections that required modifications later on and made the overall schematic easier to follow, in my opinion.
Now the schematics have been completed it was time to assign footprints and move into laying out the PCBs.
Laying out the PCBs took a little while. I constantly pushed the changes to github and quite often went back and forth between a few layouts trying to find one that worked out well enough to continue. That being said the final layout should work but is probably not the most space efficient or optimised, that will be something for version 2 maybe.
To begin I grabbed the board outlines form the CAD models from Fusion360, then imported the file into KiCAD on the edge cuts layer. I decided to start designing for the standard 2 layer boards. I went the entire way through the design working around the 2 layers, and did not see any major benefits of more layers. The track widths varied around the board depending on the power and location on the board trying to maximise the width to reduce the possibility of failure. I also put in copper fills around the board to help with the complex geometry.
I faced quite a few challenges developing the electronics design for the Tron disc, but argubly way less than the mechanical design.
Design Rules Check (DRC)
The Design Rules Check (DRC) helps check to see if the board parameters meet the specified rules. These specified rules ensure the board is manufacturable, meets the minimum spacing requirements and a few other bits too. This error occurred because of how the minimum distance is calculated and due to the angle of the parts the minimum distance calculated was less than the parameter in the DRC. To get around this i changed the DRC parameter for these parts by -0.01. This was a small enough change such that it was not an issue for manufacturing but large enough to pass the DRC.
Routing in a circle isn’t easy in KiCAD but is apparently possible in some other software packages. This meant that some of the traces are not aesthetically placed around the board. I did attempt using an open source plugin to help create some curved tracks but did not use them very much in the final design. However I did use a plugin that created circular fill regions to help create some nice round copper lines between regions. I also made use of the inbuilt shove routing once a majority of tracks were doing to push tracks into nicer locations.
Providing power to all components was a slight issue. The main problems being reaching to opposite end of the PCB. To solve this issue I used copper fills on most areas of the PCB. KiCAD connected the majority of the components. After this I went in and joined the remaining unconnected pads. This allowed me to quickly get power to the main areas of the board before routing the rest of the tracks.
The last issue was part placement. It wasn’t a major issue, more tedious. The main issues i had was trying to evenly space the parts out at the correct angles and at the right interval. I attempted some of this by hand and that’s how I did the entirety of the alight arc PCB. As for the main PCB I decided to do some research and found a plugin. The plugin allowed me to select parts and arrange them in a circle. It wasn’t perfect but at least rotated the parts to the correct angle. I then placed the parts manually lining them up with where I wanted them.
I got bored the other week and decided to design a chess set.
I decided I wanted to make the chess set quite small and travel size but didn’t want to lose the natural feel that you get from playing on a full-size board.
I conquered this by making a folding wooden chess set. All pieces and the board will be made out of wood and will be somewhat magnetic to make sure the pieces don’t go flying around in the car… I guess I might be the only one that tries to play chess in a car…
After I wrote down all of the must haves and a few wants I went to town designing. I drew up a rough sketch of what I thought the final design might look like and jumped right into Solidworks to start designing some pieces.
At the time I only had access to a few machines that could achieve the design, the main one I used was the Carvey by Inventables. This allowed me to achieve most of what I needed with a bit of hand finishing later on.
The chess pieces in Solidworks were used as blanks to ensure the right amount of room was on the inside but the final piece design was taken from another maker Beyond Design with their pieces published on the projects space run by Inventables. Here is a link the their project https://www.inventables.com/projects/chess-set
After I modelled the board in Solidworks I used an add-in to convert the parts into SVG files that the Carvey software can understand.
So with all of that here is the final product!
One last minute thing I added was a foam holder for the inside of the box. This was to stop the pieces rattling around inside, and to make sure I hadn’t lost any pieces in the car…
If you would like to make something similar let me know and I can send you through the design files.
Most digital clocks these days are made from microcontrollers, but I wanted an added challenge to use it using only logic chips trying to focus on using the TTL (74xx series) integrated circuits. So, my plan was to design and build a discrete logic clock using no microcontrollers. Since I wanted my clock to be relatively small I chose so only display Hours and Minutes, rather than hours, minutes and seconds. The clock will contain only discrete logic, no microcontrollers.
Before I designed my clock, I researched how clocks worked and explored what others had done. Most people seem to use microcontrollers, for a few reasons one being amount parts and amount of effort. The other designs I found used ripple counters. I went digging around finding out what a ripple counter was. A ripple counter is an Asynchronous (Doesn’t need a system clock) counter, that every time I get a signal on the clock pins ripples through the d-flip flops inside changing the output. each time the chip gets a pulse the ripple counter will increment the output value by one (in binary).
One part that I found common across most of them was a 32.768 kHz crystal. this seemed a perfect starting point, especially since 32768 is 2^15. in other words, if I divide the 32768 Hz signal by 2 15 times I will be left with a 1Hz signal, which is perfect because all clocks start with seconds. To divide a clock signal, I used ripple counters connected to continually divide the clock signal by 2 until the output signal was 1 Hz.
I began prototyping the 1-second segment using a CD4060. I know, not strictly TTL but this IC had a built-in oscillator amplifier and allowed me to divide by 2^14. Following this was 6 74HC93 chips which are ripple counters. these chips will increment themselves every time there is a pulse on their clock pin.
The first clock pin on the ripple counter straight after the CD4060 will be a 2Hz signal. This is because of 32768/2^14 = 2, So to get a 1 Hz signal, or in other words a tick every second you will need to divide the 1 Hz signal again by 2. The 74HC93 make this easy to do by having a separate clock pin (CP1), by feeding the output Q_0 into the CP1 pin the output on Q_1 will now be 1Hz. Now here is where things are up to you. If you want to display the second’s count you need to have an extra 74HC93 ripple counter making the total used 6. This is so you can have one ripple counter for each displayed digit.
The rest of the circuit is straightforward as the concept is very similar. The only difference is where you will connect the AND gates to reset the counters. I was making a 24-Hour clock, it would be simple to make a 12-hour clock two just by playing around with when the hour counters reset.
After all the counting the seconds, minutes and hours it might be useful to set the time. I did this by using a re-triggerable monostable multivibrator. This allowed me to press the button once and only allow the clock to increment by one. You can do this many ways, another one would be having the button link the crystal to the counter to increment it, but this would most likely always cause an overshoot. Below is the schematic for the rest of the clock.
The next step from here will be to prototype the design to make sure it works then design the circuit boards. Send me an email, or comment and let me know what you think!
It’s Finished. Check out the final product here!